What are people using in their labs to monitor and manage servers? Anyone using Spiceworks or a similar package?
What to monitor/ manage infrastructure
- Thread starter MiniKnight
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
My focus is on application monitoring, which is a bit deeper than server/network monitoring.
For application monitoring, I'm a big fan of New Relic, although the Pro pricing is far too high for home use (and probably too high for any cloud-like architecture), the lite version is merely OK, and the free version is close to useless since they trimmed off the drill-down features and keep only 30 minutes of data.
For application monitoring, I'm a big fan of New Relic, although the Pro pricing is far too high for home use (and probably too high for any cloud-like architecture), the lite version is merely OK, and the free version is close to useless since they trimmed off the drill-down features and keep only 30 minutes of data.
Last edited:
Just as a FYI - I have NR running on the main site. (Forums are having a networking challenge ATM) - My summary:
- Pro is AWESOME! I second the idea that it is too expensive for home/ cloud use. At first I thought it was $600/m on an annual contract for NR pro for 4 VMs (they do by OS not by physical machine). I was put in contact with the sales rep and she did offer a discount. Makes sense since I set up discounting structures/ deal approvals quite a bit for companies in my day job. Still - even with the discounts, the cost was going to be greater than all of the servers, spares and switches in our 2U colocation over 2 years. I have a cost analysis for the colo so fairly easy to compare.
- Lite is something I would certainly consider.
- Free is better than nothing, but certainly feels very stripped down
Here's the sad thing. If they had a $50/m plan that I could get pro monitoring for the entire 10U virtualized infrastructure I would generate $600 this year for the company.
I completely agree: They need a new pricing model that fits the new "lots of fairly small and inexpensive servers each of which isn't that important" architectures.
Looked at New Relic. Is the physical server monitoring free?
Would agree. A home or smb product at 30-50 per month would be great. Then again, these days you can use AWS and just use Amazon's monitoring tools. Not as good but something. The reason I'd pay for New Relic is because it works cross-platform.
Would agree. A home or smb product at 30-50 per month would be great. Then again, these days you can use AWS and just use Amazon's monitoring tools. Not as good but something. The reason I'd pay for New Relic is because it works cross-platform.
I'm using cacti to monitor the normal stuff exposed via SNMP and added someones IPMI script to keep an eye on the C6100 node stuff.
Don't have anything for management, only thing I've quasi stream lined is bash scripts on a Linux VM for common IPMI functions so I don't need to login to the stupid C6100 web interfaces.
Don't have anything for management, only thing I've quasi stream lined is bash scripts on a Linux VM for common IPMI functions so I don't need to login to the stupid C6100 web interfaces.
Did you post that setup? would like to see as I too would be a new relic pro customer at $5-10/mo but they price 10x higher.
I thought HP was going to do cloud management - you hook your servers into their cloud and they can use their $10B automony big data acquisition to tune/tweak/analyze your systems - imagine the amount of feedback from all HP servers - take that crunch it, generate profiles to optimize power/performance/pre-failure notification/management. While it is scarey from a hacker standpoint, you gotta admit the collective data from 1 million servers would allow massive tweak age to your servers that no one human could ever approach.
They have the tools, it makes sense. I can't wait for the day it can do all that.
They have the tools, it makes sense. I can't wait for the day it can do all that.
I haven't posted any screenshots of it. I accidentally nuked my cacti VM (was tired and deleted vhd instead of detaching it). In the process of setting it back up
Use your operating system of choice (I used Ubuntu 12.04 LTS) with cacti and Cacti ? View topic - IPMI sensors -- update June 21, 2009 for the IPMI sensors. Run a SNMP daemon on whichever machines/vms you want to monitor. Then you get the typical pretty RRD graphs.
Here's a part of a script I use to give a quick overview of what the power status is of my IPMI addressable systems:
Code:
#!/bin/sh
echo -n "c6100node: "; ipmitool -I lan -U root -P root -H nodeIP chassis status | grep "System Power" |cut -f2 -d ":"
echo -n "x9-scm-f: "; ipmitool -I lan -U ADMIN -P ADMIN -H 192.168.0.81 chassis status | grep "System Power" |cut -f2 -d ":"
Code:
alias xen='ipmitool -I lan -U root -P root -H xenserver'
Did some research:
Graphdat looks like a cool option. I can't find pricing but it says free? https://dashboard.graphdat.com/landing
Datadog has a pricing structure more what we would want to see: Datadog - Pricing -- paid version is $15/host/mo
Also New Relic Lite has a REST API: https://newrelic.com/docs/instrumentation/getting-started-with-the-new-relic-rest-api
Graphdat looks like a cool option. I can't find pricing but it says free? https://dashboard.graphdat.com/landing
Datadog has a pricing structure more what we would want to see: Datadog - Pricing -- paid version is $15/host/mo
Also New Relic Lite has a REST API: https://newrelic.com/docs/instrumentation/getting-started-with-the-new-relic-rest-api
Graphdat is free for the first server and $10/month for each additional server. I hope that their software is better than their web site - they post few details and their demo dashboard is usually broken. There is a free trial, however, so the risk is not huge.
I would be a bit worried about using their software, which is very young with just 500 total customers (including free accounts) as of March this year.
That said, I hope that the company and the product grow like gangbusters and take a big chunk of the market. With their current pricing, they could gain some serious traction if the product can deliver.
I also looked at Datadog, but I categorized them as a product that gathers and processes events but doesn't have its own agents. In other words, you'll need Datadog plus Nagios or Cacti, and/or Airbrake, etc. This is great if you are building your own devops architecture for a big infrastructure, but probably not for monitoring a few servers.
I would be a bit worried about using their software, which is very young with just 500 total customers (including free accounts) as of March this year.
That said, I hope that the company and the product grow like gangbusters and take a big chunk of the market. With their current pricing, they could gain some serious traction if the product can deliver.
I also looked at Datadog, but I categorized them as a product that gathers and processes events but doesn't have its own agents. In other words, you'll need Datadog plus Nagios or Cacti, and/or Airbrake, etc. This is great if you are building your own devops architecture for a big infrastructure, but probably not for monitoring a few servers.
Last edited:
I use Zabbix pretty heavily across our infrastructure; at one point 4 sites, nearly 500 "hosts" and 1000's of items. It's open source, and they do offer commercial support and training. Active community for help, forum and IRC channel.
Homepage of Zabbix :: An Enterprise-Class Open Source Distributed Monitoring Solution
We recently started using CopperEgg for some things, it's $$ though. Interface is buggy though.
Homepage of Zabbix :: An Enterprise-Class Open Source Distributed Monitoring Solution
We recently started using CopperEgg for some things, it's $$ though. Interface is buggy though.
Might want to also check out GroundWork. They integrate Nagios and Cacti and a number of other open source projects. Their Core product (community edition) is free for up to 50 devices.
GroundWork - The Open Platform for IT Monitoring
Overview - GroundWork
GroundWork - The Open Platform for IT Monitoring
Overview - GroundWork
+1 - If you have the time to setup Zabbix it can made a very power full monitoring tool! It has support for many enterprise features like cluster monitoring and as Swdlmarco says the community is very active.
